Chronic renal disease and stroke in atrial fibrillation: balancing the prevention of thromboembolism and bleeding risk

This editorial refers to ‘Warfarin in haemodialysis patients with atrial fibrillation: what benefit?’ by F. Yang et al., Europace 2010, 12(12):1666–1672 Atrial fibrillation (AF) confers a substantial risk of stroke and thromboembolism, and recent guidelines have advocated a risk factor-based approach to thromboprophylaxis.1 Common risk factors in everyday clinical practice have been investigated in datasets from clinical trials and cohort studies,2,3 and have been used to formulate various risk-stratification schemes, such as the CHADS2 Cardiac Failure, Hypertension, Age, Diabetes, Stroke (Doubled) score.4 Given the limitations of the latter and the non-inclusion of various ‘stroke-modifier’ risk factors,5 more comprehensive stroke-risk schemes such as the CHA2DS2-VASc Cardiac Failure, Hypertension, Age ≥75 (Doubled), Diabetes, Stroke (Doubled)—Vascular disease, Age 65–74, and Sex category (Female) score have been proposed.6 This comprehensive risk factor-based approach has been advocated in the 2010 European Society of Cardiology (ESC) guidelines on AF management.1 Even the CHA2DS2-VASc score has been (slightly) criticized for not including some risk factors associated with thromboembolism in non-valvular AF, such as cardiomyopathies (e.g. hypertrophic cardiomyopathy), infiltrative heart disease (e.g. amyloid), and renal failure. Conditions such as cardiomyopathy and amyloid are rare and while case reports and small series have suggested that such patients are at risk of stroke,7,8 they have not been specifically studied in the large randomized trials of stroke prevention in AF, to allow assessment of its predictive value for stroke on multivariate analysis. Another important consideration is the need for simplicity in any risk-stratification scheme and to have a risk score that at least can be used in everyday clinical practice, applicable to the majority of AF patients. Patients with chronic renal disease represent a complex management problem in relation to decision-making for thromboprophylaxis in AF. If these patients have severe renal impairment (e.g. creatinine clearance 65), Drugs/alcohol] score is recommended as an easy, practical way to assess bleeding risk, whereby a HAS-BLED score of ≥3 indicates ‘high-risk', and caution and/or regular review is recommended. However, ‘abnormal renal function’ scores one point on the HAS-BLED score, and many stroke risk factors are also risk factors for bleeding. Also, the HAS-BLED score has not been validated in patients with chronic kidney disease. One possible proposal is that for patients with chronic kidney disease, OAC (essentially warfarin) should be given for those at high stroke risk (e.g. a CHA2DS2-VASc score of ≥2) and that a HAS-BLED score of ≥3 would indicate regular review and follow-up is necessary. This is in contrast to the current ESC guidelines, where OAC is preferred for those with one or more stroke risk factors (i.e. CHA2DS2-VASc score of ≥1).1 A suggested schema is shown in Figure 1 (modified from19), and given the lack of large prospective clinical trial data for the new agents (e.g. dabigatran, apixaban, etc.) in severe renal impairment, we should continue to use warfarin (perhaps target INR 2.0–2.5) with careful INR monitoring, to ensure excellent anticoagulation control (with a high time in therapeutic range), and to minimize stroke and bleeding risks.20 Interestingly, dabigatran was recently approved in the United States at a dose of 75mg BID for patients with a creatinine clearance of 15–30 mL/min, despite the absence of a prospectrive controlled trial testing this dose. There would also need to be consideration of prior OAC exposure and tolerability.19 Algorithm for oral anticoagulation therapy for stroke prevention in patients with atrial fibrillation and chronic renal disease. Oral anticoagulation (OAC), for example, with Vitamin K antagonists (VKAs, target INR 2–2.5) but new drugs which may be viable alternatives to the VKAs could ultimately be considered. Asterisk (*) signifies that risk factors for stroke and thromboembolism could be assessed using the CHA2DS2-VASc score, although the real stroke risk is likely to be much higher than reported in cohorts with no renal failure. If patients have already been taking OAC (e.g. for >3 months) with no bleeding complications, these patients probably represent a patient group that would ‘tolerate’ OAC with a lower bleeding risk. Plus (+) signifies that bleeding risk could be assessed using a validated scoring system (e.g. the HAS-BLED score) although the real bleeding risk may be higher than that reported in cohorts with no renal failure. A HAS-BLED score of ≥3 indicates the need for caution and/or regular review of the patient. Such an (informal) management proposal would really need validation in multiple prospective independent cohorts, before formal incorporation into guidelines—and not beforehand. Ultimately, new OAC such as betrixaban may offer an additional option in management, pending proof from prospective outcome trials, but until then, a delicate balance will remain between stroke and thrombembolism prevention, and bleeding risk with OAC in patients with severe renal impairment. Conflict of interest: none declared.